Robust finite-time super-twisting control for magnetic field manipulation system with abrupt disturbance

Accurate measurement and analysis of physiological magnetic field signals enable early and precise diagnosis, thereby facilitating targeted interventions for cardiovascular and cerebrovascular diseases. Nevertheless, establishing an ultra-low noise environment and detecting subtle magnetic signals under abrupt disturbance remain formidable challenges. This article reformulates the dynamic model of the magnetic field manipulation system (MFMS) and proposes a robust finite-time super-twisting control (RFTSTC) strategy. A finite-time extended state observer (FTESO) is designed to estimate unmeasurable system state and disturbance, thereby enabling real-time system uncertainty quantification. Based on this, a fast super-twisting algorithm is developed on a nonsingular terminal sliding mode surface, incorporating additional linear and fractional-order terms. This operation effectively mitigates singularities, enhances convergence speed, significantly reduces oscillations, and suppresses system noise in the presence of abrupt disturbance. The finite-time convergence of the estimation errors and system states is rigorously proven using the Lyapunov method. Numerical simulations and experiment results demonstrate the superiority of the proposed strategy over existing approaches.